Pith. sign in

REVIEW 4 major objections 5 minor 54 references

Real-time decoding of quantum error correction codes using high-performance computing

T0 review · 4 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read THQLink claims that a general-purpose HPC cluster, linked to a quantum control system by a low-latency interconnect, can keep pace with real-time quantum error correction: a 2.944 µs round trip and 1 µs per syndrome round for distance-19 su

desk verdict Real systems work with a load-bearing ambiguity: the '1 µs per round' claim is decoder throughput, not closed-loop latency, and the paper never squares that with its own 2.944 µs round-trip. read the letter →

arxiv 2608.03948 v1 pith:SCX2GEVS submitted 2026-08-04 quant-ph cs.DC

Lingling Lao , Qiang Wang , Yuanqi Liu , Yantong Liu , Haowen Wang , Yitao Chen , Yankang Zhao , Zhenwei Wu
show 5 more authors
Wei Zhang Yong Dong Yingwen Liu Mingche Lai Junjie Wu
This is my paper
classification quant-phcs.DC MSC 81P70 PACS 03.67.Pp
keywords quantumerrorcorrectionreal-timedecodingsurfacecodeminimum-weightperfectmatchingparallelwindowhigh-performancecomputinginterconnectfault-tolerantlow-latencynetworking
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that a general-purpose HPC cluster, connected to a quantum processor's control system by a stripped-down, low-latency network, can satisfy the real-time decoding deadline required for fault-tolerant quantum computing. The authors build THQLink around the TH-Express interconnect, measure a mean round-trip latency of 2.944 µs between an FPGA-emulated control system and the host, and show that an overlapping parallel-window scheme running the Sparse Blossom MWPM decoder keeps up with a 1 µs-per-syndrome-round budget for surface codes up to distance 19. The significance is that real-time decoding need not live in an ASIC or FPGA: a programmable HPC cluster can both adapt to new codes and algorithms and scale out through a switched fabric that adds only about 130 ns per hop. If correct, the result says that tightly integrated QPU-HPC data centers can run software decoders in real time while also offloading hybrid quantum-classical workloads.

What carries the argument

The load-bearing mechanism is the pairing of the TH-Express interconnect with parallel-window decoding. TH-Express strips protocol overhead so FPGA-generated syndrome Mini-Packets reach host memory and corrections return in a mean 2.944 µs round trip, and its switched fabric extends decoding to more racks at about 130 ns per additional hop. The parallel-window decoder, using A/B sandwich windows with buffer regions, hides most of that transport and compute latency: A windows decode concurrently and pass seam-crossing matches as artificial defects to B windows, so throughput no longer scales with the total number of syndrome rounds. The named decoder inside each window is minimum-weight perfe

What would settle it

Run the same distance-19, 10d-round benchmark with a real quantum control system instead of the FPGA emulator, and measure t_corr − t_syn,last round by round under a continuous syndrome stream; if any round's gap exceeds the 1 µs syndrome cycle, the real-time claim is falsified.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that a general-purpose HPC cluster, rather than dedicated decoder silicon, can close the real-time QEC loop. THQLink attaches an FPGA that emulates the quantum control system to HPC racks through TH-Express, a custom interconnect with a lightweight Mini-Packet path; the measured mean round-trip latency is 2.944 µs (median 2.870 µs), with 130 ns added per switch hop. By splitting the syndrome stream into overlapping A/B windows that decode in parallel, the software MWPM decoder (Sparse Blossom) meets an amortized 1 µs-per-round budget for surface-code memory circuits up to distance 19 when 10d syndrome rounds are available at p = 0.1%; a global d

Load-bearing premise

The real-time claim depends on defining real-time as average throughput per round rather than as the single round-trip latency, and on the FPGA emulator faithfully standing in for a real quantum control system.

Editorial extensions

If this is right

  • Distance-19 surface-code memory with 10d syndrome rounds can be decoded within the 1 µs-per-round amortized budget on ordinary CPUs, so state-of-the-art matching decoders do not require custom silicon in this regime.
  • Decoding resources can scale out by adding compute racks: each extra switch hop costs about 130 ns, so expanding the fabric adds less than a microsecond of latency per hop.
  • The same HPC interconnect can absorb new QEC codes (e.g., QLDPC codes) and new decoder algorithms through software updates, avoiding the costly redesign cycles of FPGA or ASIC decoders.
  • The architecture extends naturally to tight QPU-HPC integration, where the network carries not only syndromes and corrections but also hybrid quantum-classical workloads such as error mitigation and calibration.
  • Applications that need a single-round correction before the next round begins still pay the full network round trip of about 2.944 µs, so their timing budget must account for transport, not just decode compute.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper's real-time claim is an amortized throughput statement: the 1 µs-per-round figure is met by overlapping windows, while the measured end-to-end round trip for the final syndrome round is about 2.944 µs. If a workload requires every round's correction within one cycle, the network alone leaves no slack.
  • The FPGA emulates the quantum control system rather than a real QPU control stack; a faithful test on actual control hardware could add de/serialization and correction-application latencies not captured here.
  • The 130 ns-per-hop scaling invites a distributed multi-logical-qubit design in which windows from many QEC patches are scheduled across racks; the paper does not evaluate that scheduling, leaving it as an open extension.
  • Because the fabric is general-purpose, the same interconnect could serve calibration, error mitigation, and other offloaded workloads in a quantum-centric supercomputer without adding a separate network.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes THQLink, an architecture that couples a quantum control system to an HPC cluster through the TH-Express interconnect for real-time decoding of quantum error correction codes. The authors report a mean round-trip network latency of 2.944 us, an incremental per-hop latency of 130 ns, and a parallel-window decoding strategy that, together with a CPU-based matching decoder, is claimed to achieve real-time decoding at 1 us per QEC round for surface codes up to distance 19. The paper includes measurements of network latency, end-to-end decoding latency, software-level decoding latency, and logical error rates, and compares global MWPM with the parallel-window approach.

Significance. If the central real-time claim were fully supported, this would be a useful systems contribution: it demonstrates that a programmable HPC platform with a low-latency interconnect can serve as a flexible alternative to fixed FPGA/ASIC decoders, and the reported network latency statistics are more detailed than those in comparable prior work. The paper also reports explicit sample sizes for the logical-error-rate comparison and distinguishes end-to-end latency from software-level decoding latency, which is methodologically helpful. However, the load-bearing '1 us per QEC round' claim is currently established only as decoder throughput, not as a closed-loop syndrome-to-correction latency, and the relationship between the measured 2.944 us round-trip time and the 1 us per-round budget is not resolved.

major comments (4)
  1. [Abstract, §I, §III.C, §IV.B, Fig. 10] The abstract and §I define real-time decoding as processing within about 1 us per round 'including communication latency.' In §III.C, however, the end-to-end latency tau_e2e is defined separately from software-level decoding latency, and Fig. 10 reports decoder-level (software-only) timing with syndromes already in host memory. The headline 1 us per QEC round is therefore a throughput/decoder-latency number, not a closed-loop latency. Since §IV.B reports a mean round-trip TH-Express latency of 2.944 us, a single round's syndrome-to-correction path is at least 2.944 us plus decode time, contradicting a literal reading of the abstract. The authors must either redefine the real-time criterion explicitly as amortized throughput under pipelining, or provide a timing model/experimental evidence that the closed-loop pipeline sustains 1 us per round without backlog despite the 2.944 us round tri
  2. [§IV.C, Fig. 6, Fig. 7] The text states that 'the THQLink architecture could implement real-time decoding for the distance-13 surface code' and later that parallel window 'achieves real-time decoding for distance 15' (5d rounds) and 'distance 19' (10d rounds). However, no explicit real-time threshold is given for these end-to-end latency plots. Since tau_e2e = t_corr - t_syn,last already includes a network round trip, the reader cannot tell whether the plotted values are compared against 1 us, against d us, or against total run duration. Without a stated criterion, the claim that these plots demonstrate real-time operation is not falsifiable. The authors should specify the exact deadline used and ideally overlay it on the plots.
  3. [§IV.A] The experiments use an FPGA to 'emulate the quantum control system,' but no syndrome-generation clock rate or timing fidelity characterization is reported. The real-time requirement is tied to the physical syndrome generation clock, typically about 1 us for superconducting qubits. If the FPGA does not generate and stream syndromes at the target rate with realistic round timing, the closed-loop measurements do not establish that the architecture can keep up with an actual QSC. The authors should state the emulated syndrome rate and, if possible, validate against a real control system or justify why the emulation is faithful for timing purposes.
  4. [§II.A, §IV.C] The paper itself notes in §II.A that non-Clifford logical gates such as T gates require the Pauli frame to be resolved before the gate can proceed, making latency, not only steady-state throughput, physically relevant. The reported 2.944 us round-trip latency is larger than the 1 us budget, so for T-gate-dominated circuits the architecture would miss the latency target by roughly 3x before any decoding time. If the authors intend to claim general real-time fault-tolerant decoding, they must either address this latency bound or clearly scope the claim to memory/Pauli-frame-tracking regimes where deferred correction is acceptable.
minor comments (5)
  1. [Figures 4-11] In the provided manuscript, the figure captions and axis labels appear with corrupted font encoding (e.g., '/uni00000013' sequences), making several figures unreadable. The final version must replace these with proper text.
  2. [Abstract and §IV.C] The abstract claims real-time decoding 'up to distance 19,' while §IV.C and Fig. 10 mention that parallel-window decoding reaches around 1 us at d=21 for 10d rounds and p=0.1%. Please reconcile these numbers.
  3. [§III.B] Typo: 'bulit' should be 'built.'
  4. [Figure 9/10 captions and §IV.C] The number of syndrome instances and confidence intervals are only given for Fig. 11's logical error rate. Please report sample sizes and error bars for the latency plots in Figs. 6, 7, 9, and 10 as well.
  5. [References] References [7], [51], [55], and [56] are duplicated (Skoric et al. and Tan et al. appear twice). Please consolidate.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the paper reports new measurements of a newly assembled system; the only self-citations identify TH-Express as background, not as the source of the reported numbers.

full rationale

The paper's central claims are empirical: the 2.944 µs round-trip latency, the 130 ns per-hop increment, and the decoder latencies are measured on the assembled THQLink testbed (Figs. 4, 5, 6, 7, 10). These are not predictions derived from a fitted parameter, nor are they obtained by redefining the target quantity as its input. The parallel-window decoder is benchmarked against global MWPM on identical syndrome instances (Appendix A), and its logical error rates are compared in Fig. 11, so the accuracy and latency claims are not defined into existence. The self-citations [52,53] identify TH-Express as a prior interconnect developed at the authors' institution, but they do not supply the measured round-trip or per-hop numbers; those numbers are independently measured in this paper and are externally checkable. The possible concern—that the abstract's '1 µs per QEC round' may refer to decoder-level throughput rather than the closed-loop τ_e2e that includes the 2.944 µs network round trip—is a potential correctness gap or ambiguity about the real-time criterion, not a circularity. No equation in the paper reduces to a fitted input or to a self-citation, and no load-bearing 'uniqueness theorem' is imported from the authors' prior work. The paper is self-contained as an experimental systems evaluation, so the circularity score is low.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The paper introduces a system architecture (THQLink) and uses an existing interconnect (TH-Express); no new physical entities, particles, or mediators are postulated. The only free parameters are the window geometry choices, which are hand-set and not swept. The key added assumptions are the throughput-based real-time criterion and the fidelity of the FPGA emulation of the QSC.

free parameters (1)
  • Parallel-window step size s and buffer size b = (d+1)/2 each
    Chosen by hand, not swept. The window span w = s + 2b = 3(d+1)/2 rounds and the resulting parallelism directly determine the reported 1 µs/round throughput; other choices would change all latency results.
assumptions (4)
  • domain assumption Circuit-level depolarizing noise model with per-gate/reset/measurement error probability p
    Adopted in §IV.A; all latency and logical-error results are conditioned on this noise model, which may not match real device noise.
  • ad hoc to paper The FPGA emulating the QSC faithfully reproduces the timing behavior and interface of a real quantum control system
    §IV.A states the FPGA emulates the quantum control system. The real-time conclusions depend on this emulation fidelity, which is not validated against any actual QPU.
  • domain assumption The parallel-window A/B decomposition preserves global matching information sufficiently for fault tolerance
    Imported from Tan et al. [55] and Skoric et al. [56]. The paper verifies only a slightly higher P_fail at p=0.1% for d×d×5d up to d=15, not for the larger d and 10d cases where the real-time claims are made.
  • ad hoc to paper Throughput of 1 µs per QEC round, rather than single-round end-to-end latency, is the binding real-time constraint
    The paper's τ_e2e includes a 2.944 µs network round trip yet the headline budget is 1 µs; the real-time claim rests on the assumption that amortized throughput is the correct criterion. This is not reconciled with Section I's statement that the real-time budget includes communication latency.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Real-time decoding of quantum error correction codes using high-performance computing." pith.science (2026). https://pith.science/paper/SCX2GEVS

@misc{pith2026260803948,
  author       = {Pith},
  title        = {Pith review of: Real-time decoding of quantum error correction codes using high-performance computing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SCX2GEVS}},
  note         = {Machine review of arXiv:2608.03948}
}
abstract

Quantum error correction (QEC) is indispensable for building scalable fault-tolerant quantum computers. Effective QEC demands stringent real-time decoding: the decoder must process syndrome measurements and determine corrections within a time scale--typically on the order of microseconds, to avoid data backlog. Scaling to large number of logical qubits further necessitates significant computational resources. In this work, we propose an architecture, called \emph{THQLink}, for real-time decoding of quantum error correction codes using high-performance computing (HPC) resources. The network connecting the HPC and the control system of quantum processing unit (QPU) is built on TH-Express and can be adapted to different quantum technologies and their associated control stacks. We report a round-trip latency of 2.944 $\mu$s on average, with an incremental overhead of 130 ns per additional hop. Using a parallel window strategy, we demonstrate real-time decoding (1 $\mu$s per QEC round) of the surface code up to distance 19 using a matching-based decoder on CPUs. Our work presents a scalable framework for real-time decoding in fault-tolerant quantum computing. It can be readily applied to quantum-centric supercomputers that feature tight integration between QPU and HPC resources, thereby enabling efficient support for hybrid quantum-classical algorithms and computation-intensive workloads offloaded from the QPU.

Figures

Figures reproduced from arXiv: 2608.03948 by the authors.

Figure 1
Figure 1. FIG. 1. The rotated surface code and its decoding graphs. (a) [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Overview of the [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. The real-time decoding flow via the high-speed network [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: FIG. 4. The round latency of TH-Express with direct inter [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Latency overhead per additional hop [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: reports the end-to-end latency for decoding d rounds of syndromes (i.e., d × d × d) at physical er￾ror rates p = 0.1% and p = 0.3% using global MWPM. For physical error rate p = 0.1%, the THQLink architec￾ture could implement real-time decoding for the distance￾13 surf…
Figure 7
Figure 7. Figure 7: FIG. 7. End-to-end decoding latency for [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8. The parallel window scheme [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 9
Figure 9. Figure 9: FIG. 9. Software-level decoding latency per round for [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: FIG. 10. Decoder-level latency for [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]
Figure 11
Figure 11. Figure 11: FIG. 11. Logical error rate [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

54 extracted references · 24 canonical work pages

  1. [1]

    P. W. Shor, Physical review A52, R2493 (1995)

    work page 1995

  2. [2]

    A. M. Steane, Physical Review Letters77, 793 (1996)

    1996

  3. [3]

    B. M. Terhal, Reviews of Modern Physics87, 307 (2015)

    2015

  4. [4]

    Gottesman,Stabilizer codes and quantum error cor- rection(California Institute of Technology, 1997)

    D. Gottesman,Stabilizer codes and quantum error cor- rection(California Institute of Technology, 1997)

    1997

  5. [6]

    A. G. Fowler, M. Mariantoni, J. M. Martinis, and A. N. Cleland, Phys. Rev. A86, 032324 (2012)

    2012

  6. [7]

    Skoric, D

    L. Skoric, D. E. Browne, K. M. Barnes, N. I. Gillespie, and E. T. Campbell, Nature Communications14, 7040 (2023)

    2023

  7. [8]

    G. Q. AI and Collaborators, Nature638, 920 (2025)

    2025

  8. [9]

    Caune, L

    L. Caune, L. Skoric, N. S. Blunt, A. Ruban, J. McDaniel, J. A. Valery, A. D. Patterson, A. V. Gramolin, J. Ma- janiemi, K. M. Barnes,et al., Nature Communications (2026)

    work page 2026

Show all 54 references
  1. [10]

    Dennis, A

    E. Dennis, A. Kitaev, A. Landahl, and J. Preskill, Jour- nal of Mathematical Physics43, 4452 (2002)

    2002

  2. [11]

    Higgott, ACM Transactions on Quantum Computing 3, 1 (2022)

    O. Higgott, ACM Transactions on Quantum Computing 3, 1 (2022)

    2022

  3. [12]

    Higgott and C

    O. Higgott and C. Gidney, Quantum9, 1600 (2025)

    2025

  4. [13]

    Wu and L

    Y. Wu and L. Zhong, in2023 IEEE International Con- ference on Quantum Computing and Engineering (QCE), Vol. 1 (IEEE, 2023) pp. 928–938

    2023

  5. [14]

    Delfosse and N

    N. Delfosse and N. H. Nickerson, Quantum5, 595 (2021)

    2021

  6. [15]

    Poulin and Y

    D. Poulin and Y. Chung, arXiv preprint arXiv:0801.1241 (2008)

    2008 arXiv

  7. [16]

    M¨ uller, T

    T. M¨ uller, T. Alexander, M. E. Beverland, M. B¨ uhler, B. R. Johnson, T. Maurer, and D. Vandeth, arXiv preprint arXiv:2506.01779 (2025)

    2025 arXiv

  8. [17]

    J. Olle, R. Zen, M. Puviani, and F. Marquardt, npj Quan- tum Information10, 136 (2024)

    2024

  9. [18]

    Sweke, M

    R. Sweke, M. S. Kesselring, E. P. van Nieuwenburg, and J. Eisert, Machine Learning: Science and Technology2, 025005 (2021)

    2021

  10. [19]

    E. S. Matekole, E. Ye, R. Iyer, and S. Y.-C. Chen, arXiv preprint arXiv:2212.11890 (2022)

    2022 arXiv

  11. [20]

    A. Y. He and Z.-W. Liu, arXiv preprint arXiv:2502.14372 (2025)

    2025 arXiv

  12. [21]

    Lange, P

    M. Lange, P. Havstr¨ om, B. Srivastava, I. Bengtsson, V. Bergentall, K. Hammar, O. Heuts, E. van Nieuwen- burg, and M. Granath, Physical Review Research7, 023181 (2025)

    2025

  13. [22]

    A. Gong, S. Cammerer, and J. M. Renes, in2024 IEEE International Symposium on Information Theory (ISIT) (IEEE, 2024) pp. 2700–2705

    2024

  14. [23]

    A. S. Maan and A. Paler, npj Quantum Information11, 78 (2025)

    2025

  15. [24]

    Z. Liu, A. Gong, and B. K. Clark, arXiv preprint arXiv:2509.22347 (2025)

    2025

  16. [25]

    H. Cao, F. Pan, D. Feng, Y. Wang, and P. Zhang, arXiv preprint arXiv:2503.21374 (2025)

    2025 arXiv

  17. [26]

    Bausch, A

    J. Bausch, A. W. Senior, F. J. Heras, T. Edlich, A. Davies, M. Newman, C. Jones, K. Satzinger, M. Y. Niu, S. Blackwell,et al., Nature635, 834 (2024)

    2024

  18. [27]

    A. W. Senior, T. Edlich, F. J. Heras, L. M. Zhang, O. Higgott, J. S. Spencer, T. Applebaum, S. Black- well, J. Ledford, A. ˇZemgulyt˙ e,et al., arXiv preprint arXiv:2512.07737 (2025)

    2025

  19. [28]

    Zhang, Z

    K. Zhang, Z. Yi, S. Guo, L. Kong, S. Wang, X. Zhan, T. He, W. Lin, T. Jiang, D. Gao,et al., arXiv preprint arXiv:2601.09921 (2026)

    2026

  20. [29]

    Choukroun and L

    Y. Choukroun and L. Wolf, inAdvances in Neural In- formation Processing Systems (NeurIPS), Vol. 35 (2022) pp. 23390–23402

    2022

  21. [30]

    Cohen, Y

    S.-e. Cohen, Y. Choukroun, and E. Nachmani, arXiv preprint arXiv:2505.17834 (2025)

    2025 arXiv

  22. [31]

    Barber, K

    B. Barber, K. M. Barnes, T. Bialas, O. Bu˘ gdaycı, E. T. Campbell, N. I. Gillespie, K. Johar, R. Rajan, A. W. Richardson, L. Skoric,et al., Nature Electronics8, 84 (2025)

    2025

  23. [32]

    Maurya, T

    S. Maurya, T. Maurer, M. B¨ uhler, D. Vandeth, and M. E. Beverland, arXiv preprint arXiv:2511.21660 (2025)

    2025

  24. [33]

    Maurer, M

    T. Maurer, M. B¨ uhler, M. Kr¨ oner, F. Haverkamp, T. M¨ uller, D. Vandeth, and B. R. Johnson, arXiv preprint arXiv:2510.21600 (2025)

    2025

  25. [34]

    A. B. Ziad, A. Zalawadiya, C. Topal, J. Camps, G. P. Geh´ er, M. P. Stafford, and M. L. Turner, Nature Com- munications16, 11048 (2025)

    2025

  26. [35]

    Y. Wu, N. Liyanage, and L. Zhong, inProceedings of the 30th ACM International Conference on Architectural Support for Programming Languages and Operating Sys- tems, Volume 2(2025) pp. 639–654

    2025

  27. [36]

    P. Das, C. A. Pattison, S. Manne, D. Carmean, K. Svore, M. Qureshi, and N. Delfosse, arXiv preprint arXiv:2001.06598 (2020)

    2001 arXiv

  28. [37]

    S. A. Caldwell, M. Khazraee, E. Agostini, T. Lassiter, C. Simpson, O. Kahalon, M. Kanuri, J.-S. Kim, S. Stan- wyck, M. Li,et al., arXiv preprint arXiv:2510.25213 (2025)

    2025

  29. [38]

    M. A. Nielsen and I. L. Chuang,Quantum computation and quantum information(Cambridge university press, 2010)

    2010

  30. [39]

    P. W. Shor, inProceedings of 37th conference on founda- tions of computer science(IEEE, 1996) pp. 56–65

    1996

  31. [40]

    A. R. Calderbank and P. W. Shor, Physical Review A 54, 1098 (1996)

    1996

  32. [41]

    A. Y. Kitaev, Annals of physics303, 2 (2003)

    2003

  33. [42]

    S. B. Bravyi and A. Y. Kitaev, quant-ph/9811052 (1998)

    1998 arXiv

  34. [43]

    Bombin and M

    H. Bombin and M. A. Martin-Delgado, Phys. Rev. Lett. 97, 180501 (2006)

    2006

  35. [44]

    Kubica and M

    A. Kubica and M. E. Beverland, Physical Review A91, 032330 (2015)

    2015

  36. [45]

    A. M. Kubica,The ABCs of the color code: A study of topological quantum codes as toy models for fault- tolerant quantum computation and quantum phases of matter(California Institute of Technology, 2018)

    2018

  37. [46]

    Bomb ´ ın, New Journal of Physics17, 083002 (2015)

    H. Bomb ´ ın, New Journal of Physics17, 083002 (2015)

    2015

  38. [47]

    Bravyi, A

    S. Bravyi, A. W. Cross, J. M. Gambetta, D. Maslov, P. Rall, and T. J. Yoder, Nature627, 778 (2024)

    2024

  39. [48]

    C. Zhao, C. Duckering, A. Gu, N. Maskara, and H. Zhou, arXiv preprint arXiv:2604.16209 (2026)

    2026 arXiv

  40. [49]

    Bluvstein, A

    D. Bluvstein, A. A. Geim, S. H. Li, S. J. Evered, J. P. 9 Bonilla Ataides, G. Baranes, A. Gu, T. Manovitz, M. Xu, M. Kalinowski,et al., Nature649, 39 (2026)

    2026

  41. [50]

    Paetznick, B

    A. Paetznick, B. Reichardt, M. da Silva, C. Ryan- Anderson, D. Aasen, J. Bello-Rivas, J. Campora, R. Chao, A. Chernoguzov, W. van Dam,et al., Nature 654, 349 (2026)

    2026

  42. [52]

    S. Xu, M. Lai, Y. Dai, J. Cao, and K. Wang, Neurocom- puting393, 126 (2020)

    2020

  43. [53]

    M. Xie, Y. Lu, K. Wang, L. Liu, H. Cao,et al., IEEE micro32, 8 (2011)

    2011

  44. [54]

    P. Das, A. Locharla, and C. Jones, Proceedings of the 27th ACM International Conference on Architectural Support for Programming Languages and Operating Sys- tems , 541 (2022)

    2022

  45. [55]

    X. Tan, F. Zhang, R. Chao, Y. Shi, and J. Chen, PRX Quantum4, 040344 (2023)

    2023

  46. [56]

    Skoric, D

    L. Skoric, D. E. Browne, K. M. Barnes, N. I. Gillespie, and E. T. Campbell, Nature Communications14, 7040 (2023). Appendix A: Parallel window decoding For fault-tolerant quantum error correction, the de- coder will process syndromes over many rounds. Batch (global) decoders ha...

    2023

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.